Quantification of plasma lipids, lipoproteins, and apolipoproteins is often required to screen apparently healthy individuals and for followup of patients with dyslipoproteinemias. Low-density lipoprotein (LDL) cholesterol is likely the most atherogenic lipoprotein cholesterol fraction, and high-density lipoprotein (HDL) cholesterol is known to be strongly antiatherogenic in humans. We studied the performance of several new methods for measuring LDL cholesterol and HDL cholesterol in clear and lipemic human sera. Besides calculated LDL cholesterol (Friedewald formula), three new "direct" LDL cholesterol techniques were assessed. The Sigma direct LDL cholesterol method is based on removal of HDL and very low- density lipoprotein (VLDL) by immunoabsorption and measurement of total cholesterol in the absorbed serum. The Polymedco method removes LDL and calculates LDL cholesterol as the difference between total cholesterol and the cholesterol content of the remaining lipoproteins. The Helena LipidProfile cholesterol method involves quantitative assessment of cholesterol by a coupled enzyme reaction after electrophoretic separation of various lipoprotein fractions in agarose gel. Hence, this method concomitantly measures cholesterol in all major lipoprotein fractions such as HDL, lipoprotein(a), VLDL, LDL, and chylomicrons. In addition to determination with lipoprotein cholesterol electrophoresis, HDL cholesterol was also determined by the classic dextran sulfate-Mg2+ precipitation method and by a new, solid-phase dextran sulfate-Mg2+ method (Polymedco) where the non-HDL is magnetically removed by means of dextran sulfate-coated particles from the serum. While all lipoprotein cholesterol methods were capable of providing analytically acceptable results in clear human sera, only the electrophoretic method for both LDL cholesterol and HDL cholesterol, the Sigma method for LDL cholesterol, and the Polymedco method for HDL cholesterol gave consistently reliable measurements in lipemic sera. Thus at present only these three routine methods are expected to provide diagnostically reliable results in cases of persistent lipemia or non fasting specimens.